Tech evaluation of the Iron Driver Loudspeakers

[JRT]

Re: Tech evaluation of the Iron Driver Loudspeakers

Ideally we should have blindfolds or dark shades to eliminate the influence of the screen. The point was valid and may be for the next competition, shades would be an option.

Maybe get some of those extra large ultra-cheap disposabe safety glasses and spray the front with BIN primer, zinc oxide loaded shellac.

[Jeff B.]

Re: Tech evaluation of the Iron Driver Loudspeakers

[ATTACH=CONFIG]29683[/ATTACH]
[ATTACH=CONFIG]29684[/ATTACH]


Thanks to all,
David

You're not saying you should be designing based on Fletcher-Munsen curves, are you?

[David Crewe]

Re: Tech evaluation of the Iron Driver Loudspeakers

Certainly the FR should not be based on the curves, maybe governed was to strong a word. I only offer them up as a possable explination for why I seem to prefer a FR that is a bit relaxed between 1k and 6k and has a rise on the top end. It seems that a great many seem to shoot for dead flat with a top end that is flat or falling off which for me tends to suck the life out music. I actualy took about 1 1/2 db off the top end just before boxing them up if fear that they may considered to bright.
Maybe it's just me but, it seems to work.
David

[David Crewe]

Re: Tech evaluation of the Iron Driver Loudspeakers

Certainly the FR should not be based on the curves, maybe governed was to strong a word. I only offer them up as a possable explination for why I seem to prefer a FR that is a bit relaxed between 1k and 6k and has a rise on the top end. It seems that a great many seem to shoot for dead flat with a top end that is flat or falling off which for me tends to suck the life out music. I actualy took about 1 1/2 db off the top end just before boxing them up if fear that they may considered to bright.
Maybe it's just me but, it seems to work.
David

[mzisserson]

Re: Tech evaluation of the Iron Driver Loudspeakers

You're not saying you should be designing based on Fletcher-Munsen curves, are you?

What is wrong with it? I know I have been asking a lot of unanswered questions lately, but I would like to know, in a non nebulous way, why consideration of how we hear to designing a device that mates to it is such a bad thing?

Certainly the FR should not be based on the curves, maybe governed was to strong a word. I only offer them up as a possable explination for why I seem to prefer a FR that is a bit relaxed between 1k and 6k and has a rise on the top end. It seems that a great many seem to shoot for dead flat with a top end that is flat or falling off which for me tends to suck the life out music. I actualy took about 1 1/2 db off the top end just before boxing them up if fear that they may considered to bright.
Maybe it's just me but, it seems to work.
David

Dave you really hit it here. It was proven with the ONLY design I ever made flat. It placed 4th. I like how you are looking at the total design of the loudspeaker and considering its end user. You woudl not mate a molex B connector to a D-sub, would you? You can adapt it (EQ), but the right connector to the right connector is always ideal. I have gotten crucified in the past for implying what you have stated above, and I think the data speaks for itself in many ways. Its funny, Roman said to me he had the winning speaker when he got yours in. (WE DID NOT JUDGE SO HERE IS ZERO INFLUENCE) I was excited to hear Roman that floored by any speaker as he is a man of method and not surprise. Then when I heard them I was pretty blown away. Remember, you only lost best sound by 0.25 points, it was the closest score of any two speakers, one point on one sheet and it was tied, or tipped in your favor. I think it begins to paint a picture of design methodology where there is no one measurement, or school of thought that can dictate good design, regardless of how quantified everyone tries to make it and concideration of how WE work, not just the speakers is indeed an important one.

[r-carpenter]

Re: Tech evaluation of the Iron Driver Loudspeakers

Thanks for the measurements Roman. A fair amount of ringing there. I wonder if Monoiz is able to get in there after the fact and brace it up better. It's his pair of speakers so it's his choice.

I was actually revisiting "Measuring Loudspeakers" by D'Appolito today and he writes that ringing panels will produce ridges in CSD but the wavelength of the resonant F is equal to the size of the panel. A standing wave according to D'Appolito will equal half of the distance between two surfaces. So, based on his book, your panels are ringing and it's not a padding and staffing issue. I've seen standing waves equal to the internal dimensions of the loudspeaker case, so probably Matt may want to look in to his case a bit deeper, if there's interest. D'Appolito also talks about ridges on the impedance curve which should have shown on the impedance plot but I've never seen these with Sound Easy, LIMP, WT2 or DATS. May be MLSSA is a way to go for better set up.

[ROTECH]

Re: Tech evaluation of the Iron Driver Loudspeakers

Dave you really hit it here. It was proven with the ONLY design I ever made flat. It placed 4th. I like how you are looking at the total design of the loudspeaker and considering its end user. You woudl not mate a molex B connector to a D-sub, would you? You can adapt it (EQ), but the right connector to the right connector is always ideal. I have gotten crucified in the past for implying what you have stated above, and I think the data speaks for itself in many ways. Its funny, Roman said to me he had the winning speaker when he got yours in. (WE DID NOT JUDGE SO HERE IS ZERO INFLUENCE) I was excited to hear Roman that floored by any speaker as he is a man of method and not surprise. Then when I heard them I was pretty blown away. Remember, you only lost best sound by 0.25 points, it was the closest score of any two speakers, one point on one sheet and it was tied, or tipped in your favor. I think it begins to paint a picture of design methodology where there is no one measurement, or school of thought that can dictate good design, regardless of how quantified everyone tries to make it and concideration of how WE work, not just the speakers is indeed an important one.

Can I get an Amen. Very well said. People are as important to speaker design as speaker design is to people.
This competition may have had its opposition based on the driver used being so easy to work with, but that lends itself to the true meaning of a design competition and not a "choice of drivers" competition.

[ryanbouma]

Re: Tech evaluation of the Iron Driver Loudspeakers

What is wrong with it? I know I have been asking a lot of unanswered questions lately, but I would like to know, in a non nebulous way, why consideration of how we hear to designing a device that mates to it is such a bad thing?

IMO, it's a bad thing because we are reproducing what the artists and recording engineers are producing, and their ears are already subject to the fletcher munson curves. For instance, when I played guitar, I knew how lightly or agressively I wanted certain notes to sound when in where in the song. Because it sounds good. The role of the speaker is to replay it. If the speaker adds aggressiveness to a certain band of frequencies, then it'll sound too aggressive. Or it'll sound agressive when it's meant to sound subtle.

The same thing comes up with hearing damage. I often hear people say, well, I like tilted up treble because I have bad hearing. It's my belief that the brain compensates for our hearing damage over time, within reason. We know what human speach is supposed to sound like. If the speaker reproduces it with a varied frequency response, it's sound wrong, damaged hearing or not. If you have bad hearing, does it mean when I talk to you I'll sound wrong and that I should eq my voice to sound correct?

IMO flat is correct. There are exceptions to this. For instance just back a couple pages I said I don't think flat is correct for the iron driver in the upper bass lower mids (Baffle step region) because of the drivers inherent sound. Or if we have a high dispersion design (omni maybe) and we want to focus on power response rather than on axis response. Etc. But if we have ideal drivers, flat is correct. Ideal drivers is something different than "how we hear".

Also important to add, if someone enjoys a speaker with fat bass and rolled off treble, so be it. That's their choice and all the power to them to enjoy that sound.

EDIT - This is a useful white paper: http://www.harman.com/EN-US/OurCompa...ndRoomsPt2.pdf

It is white and subject to commercial influence from the Harmon group, but from my understanding it aligns with Toole's book Sound Reproduction which is an esteem plublication afaik. The white paper is a quick synopsis of his book (which I have yet to read) and is a good look into what we're trying to achieve. Again, all IMO (in my opinion).

[ryanbouma]

Re: Tech evaluation of the Iron Driver Loudspeakers

"Measuring Loudspeakers" by D'Appolito today and he writes that ringing panels will produce ridges in CSD but the wavelength of the resonant F is equal to the size of the panel.

Thanks. Which document is this? I have the book Testing Loudspeakers by D'Appolito. And I have a copy of his Audio Express article. Where can I read about this? Thanks.

[jeff_free69]

Re: Tech evaluation of the Iron Driver Loudspeakers

I seem to remember grading ours high on treble quality, but low on mid/bass balance because in the room it seemed to have less than most of the designs.
In my bedroom the bass is huge ...

Yes, very interesting, that's the exact reason I voted for them!

I am uniquely advantaged, and at the same time, disadvantaged since I'm so intimately familiar with the source material on the CD. While I know the nuances and pros and cons of each track, I heard them many dozens of times on my own speakers, so that undoubtedly had an influence (For reference, the speakers I use are Zaph SR71 (plus a sub) in my home studio, Statement Monitors in the living room and Overnight Sensations on the computer. )

To me the first distinguishing factor was the "emotional involvement" which to me correlates to mid/treble presence, smoothness and clarity. Several had that, so the next differentiator was getting the bass just right. At my location up front, most of the speakers presented more bass than I was expecting. But we may attribute that to the room - If I leaned forward it was more exagerated.

For the record I ended up picking yours and the "Saffrons" in a tie for first. (without side by side comparisons that's as good as I can do)
If I'm getting all the graphs straight, and looking at the "merged", both had a rise in the high end. Once again, my position being up front and center, I was more off axis than most, so that would even out the rise ( i believe the speakers were not toed-in, but please correct me if i'm wrong). plus the curtain dampened things an extra db ( I guess everybody's designing in an extra db next year

All that being said, I'm relieved my perceptions were still in step with people in the back.


*************************
Great job everyone!

I think this raised the bar on DIY events on several fronts:

- Blind listening. I'm sure I'd have been swayed by the looks

- A common set of measurements. Great work by Roman to get this done. And I'm really impressed by the fact that everyone went the extra mile to ship the speakers ahead of time.

- A built-in Beer tasting

[Brewski]

Re: Tech evaluation of the Iron Driver Loudspeakers

Jeff I think it was the last track in the demo that really defined how I liked a speaker's bass output I changed a couple scores due to that track since bass that was impressively low in some instances seemed to end up showing poor control during that track and get a bit muddy in some cases.

Take it easy
Jay

[mzisserson]

Re: Tech evaluation of the Iron Driver Loudspeakers

IMO, it's a bad thing because we are reproducing what the artists and recording engineers are producing, and their ears are already subject to the fletcher munson curves. For instance, when I played guitar, I knew how lightly or agressively I wanted certain notes to sound when in where in the song. Because it sounds good. The role of the speaker is to replay it. If the speaker adds aggressiveness to a certain band of frequencies, then it'll sound too aggressive. Or it'll sound agressive when it's meant to sound subtle.

The same thing comes up with hearing damage. I often hear people say, well, I like tilted up treble because I have bad hearing. It's my belief that the brain compensates for our hearing damage over time, within reason. We know what human speach is supposed to sound like. If the speaker reproduces it with a varied frequency response, it's sound wrong, damaged hearing or not. If you have bad hearing, does it mean when I talk to you I'll sound wrong and that I should eq my voice to sound correct?

IMO flat is correct. There are exceptions to this. For instance just back a couple pages I said I don't think flat is correct for the iron driver in the upper bass lower mids (Baffle step region) because of the drivers inherent sound. Or if we have a high dispersion design (omni maybe) and we want to focus on power response rather than on axis response. Etc. But if we have ideal drivers, flat is correct. Ideal drivers is something different than "how we hear".

Also important to add, if someone enjoys a speaker with fat bass and rolled off treble, so be it. That's their choice and all the power to them to enjoy that sound.

EDIT - This is a useful white paper: http://www.harman.com/EN-US/OurCompa...ndRoomsPt2.pdf

It is white and subject to commercial influence from the Harmon group, but from my understanding it aligns with Toole's book Sound Reproduction which is an esteem plublication afaik. The white paper is a quick synopsis of his book (which I have yet to read) and is a good look into what we're trying to achieve. Again, all IMO (in my opinion).

I know you have not read toole's book. He talks about the difference between a micorphone and an ear and how fal on axis responce can contribute to a speaker that can be precieved wose than one whic exibhits slight tonal characteristics that mimic the FM curve. You you say flat is correct, but you design your speakers otherwise...I don't get it?

That is a great paper by Harmon, thanks for the link.

[ryanbouma]

Re: Tech evaluation of the Iron Driver Loudspeakers

I know you have not read toole's book. He talks about the difference between a micorphone and an ear and how fal on axis responce can contribute to a speaker that can be precieved wose than one whic exibhits slight tonal characteristics that mimic the FM curve.

Interesting and news to me. I haven't heard of this find before. None of Toole's work, including the white paper I linked, that I'm familiar with indicates this. I'm interested in knowing more about it.

You you say flat is correct, but you design your speakers otherwise...I don't get it?

Like I explained. Some drivers require compensation. Essentially, I feel the iron driver is not ideal (no driver is) in the lower mids/upper bass and sounds extra full. So it's appropriate to compensate for that. I'm not compensating for a sound preference or the FM curves, I'm compensating for driver behaviour. Just like, if you were planning on listening off axis to a 4" full range driver, you would likely boost the treble because it beams. It's not an ideal driver and measures of compensation are required.

That is a great paper by Harmon, thanks for the link.

You're welcome.

[Jeff B.]

Re: Tech evaluation of the Iron Driver Loudspeakers

What is wrong with it? I know I have been asking a lot of unanswered questions lately, but I would like to know, in a non nebulous way, why consideration of how we hear to designing a device that mates to it is such a bad thing?

It appears that Ryan pretty much saved me from posting, as he already covered most of what would have been my main points. So, I merely clarify some of it in my own words.

First, of all let's take a look at those FM curves, and reference the line in the 80-90 Phons range. Surely no one here would ever advocate putting a 10-15 dB dip in their speaker's frequency response at 4khz, or add a 10dB peak at 10khz, or heaven forbid a boost of 30dB or more in the bass (however, I believe there are some car stereo around here that do just that ;-). So, that means no one would actually advocate designing to these curves, and if they did their speakers would not sound very good to anyone, them included.

The fact is, these curves have nothing whatsoever to do with speaker design (with a very small exception, which I will comment on at the end), they have to do with our natural hearing sensitivity for what is called "equal loudness" vs frequency vs SPL. In other words the bass must be 30dB louder than the upper midrange before we interpret that sound as equally loud, and so on. But what seems to get lost in the translation is in the fact that we hear everything according to these curves - people talking to us, a baby crying, a jazz ensemble playing, a full orchestra, a jet taking off, etc. These curves are already factored in to how we hear everything in nature.

When you listen to Joni Mitchell singing to an acoustic guitar live you are perceiving according to these curves already. So, if this session was recorded and played back for you would it make any sense to apply these curves to it again? If you do, the playback will certainly sound nothing at all like the original performance. However, if it is played back reproduced as closely as possible to the original, then it will sound as closely as possible to the original to your ears - because, as I said, these curves are already present in the way you are hearing everything anyway. You don't want all sounds to be equally loud, and that's what these curves are about in the first place.

Now, I am not really talking about whether a speaker will sound best if it is designed "flat" or not, because someone will first need to define what "flat" even means. "Flat" is a very cloudy term to begin with - flat on what axis? there are how many of them now? Flat at what distance? Flat in what room, an anechoic chamber, or your living room? and it goes on and on. It is because of these reasons that small changes in frequency response, what we might call "voicing", becomes an important part of the design process. And I am sure this is really what you are talking about and not about designing to FM curves, right?

My Continuum, for example, has some very specific and intentional deviations from what someone might call flat (even though they all fall within +/- 2.5db across the board), but most everyone comments on how natually voiced and balanced they sound, especially with voices. This because these deviations in the on-axis response are specific to how the off-axis response combines with it and how the room tends to impact the response, etc., (there's really a lot of thought and research that went into this) so in the end, the perceived sound seems "correct" to our ear.

I suspect that when all of the semantics get ironed out we will be much closer on all of this than you thought.

Jeff B.

[jsr]

Re: Tech evaluation of the Iron Driver Loudspeakers

Wow, great speakers, great measurements and great info (I was wondering about how those FM curves factored in recently).

Going back to the "Iron Driver", I had asked for opinions on the 830656 a short while back and the feedback was a fairly cohesive "meh"...not great, not poor...it's decent and easy to work with. Seeing the various implementations in their final forms, is that still everyone's opinion of the 830656?

Thanks.